A NEW T-SHAPED GRAPHITE FURNACE FOR ATOMIC … · providing us the atomic absorption spectrometer...
Transcript of A NEW T-SHAPED GRAPHITE FURNACE FOR ATOMIC … · providing us the atomic absorption spectrometer...
-
A NEW T-SHAPED GRAPHITE FURNACE FOR
ATOMIC ABSORPTION SPECTROMETRY
Vom Fachbereich Chemie (IAC)
der Universität Duisburg-Essen
Zur Erlangung des akademischen Grades eines
Dr. rer. nat
genehmigte Dissertation
von
Abdelsalam Ali Asweisi
aus Benghazi / Libyen
Referent: Prof. Dr. Heinz-Martin Kuss
Korreferent: Priv. Doz. Dr. Ursula Telgheder
Datum der Einreichung 02.11.2007
Datum der mündlischen prüfung 07.02.2008
-
AKNOWLEDGEMENTS
During the work on this thesis, I have been lucky to work in a research group that has helped
me a lot and made this really very pleasant, I would like to thank them all, but particularly my
thesis director Prof. Dr. Heinz-Martin Kuss, who has been constantly by my side, teaching,
advising, discussing, supporting and easing the moments of indecision, thank you Prof. Kuss
for guiding me in the first steps of my research project.
I must not forget my first colleagues, Dr.Bülend Bayraktar, who showed me to the basics of
graphite furnace technique and Dr. Yevgen Berkhoyer, who helped me to start out in research
and introduced me to the method of two-step atomizer.
I would like to thank Mr. Theo Lukaszyk for paying more attention during the manufacturing
of the graphite furnaces and the rest of all parts which were needed during my research period
and Mr. Jürgen Kupperschmied the leader of the University fine mechanic workshop.
I express my sincere gratitude to Dr. Uwe Oppermann from Shimadzu company-Germany for
providing us the atomic absorption spectrometer (AA6800).
Schunk kohlenstofftechnik Gmbh, thanks for providing graphite materials and pyrocoating
processes.
-
I especially want to thank Priv. Doz. Dr. Ursula Telgheder for here significant suggestions
and final corrections of my thesis and to do the second reference.
I would like to acknowledge Dr. Holger Krohn, Dr. Bernd Wermeckes, Mr. Gerd Fischer, Mr.
Werner Kaiser, Mrs. Roswitha Schragmann, Mrs. Claudia Ullrich for their assistance during
my research period.
My friends Rajab El-kailany, Nabil Bader, Kahled Elsherif, Dr. Roman Rodreguez, Stefan
Meisen, Sami Abdelsalam and Qian Yuan are warmly acknowledged.
I would like to acknowledge Mr. Mohamed Altous for helping me in drawing graphs using
three dimensional software program.
I want also thank Prof. Trosten Schmidt, Prof. Karl Molt, Prof. Alfred Golloch, Dr. Myint
Sein for the friendly atmosphere during my work in this department. All co-workers in
instrumental analysis institute are also acknowledged.
Finally, my warm thanks to my wife for taking care of me and our kids and her patience
during my study period.
-
V
TABLE OF CONTENT
1. INTRODUCTION. ………………………………………………………………. 1
1.1 Trace element analysis. …………………………………………... ………… 1
1.2 Spectrometric methods. ……………………………………………………… 1
1.3 Atomic absorption spectrometry (AAS). …………………………………… 1
1.3.1 Instrumentation. ………………………………………………………. 1
1.3.2 Measuring Absorbance. ……………………………………………… 2
1.3.3 THE graphite furnace. ………………………………………………. 4
1.3.3.1 Short history of graphite tube. ……………………………….. 4
1.3.3.2 STPF concept …………………………………………………. 5
1.3.4 The temperature program. ………………………………………… 6
1.4 Interferences in atomic absorption spectrometry. ……………………. 7
1.5 Chemical modifiers. …………………………………………………….. 8
2. THEORETICAL BACKGROUND KNOWLEDGE. ……………………… 11
2.1 Types of graphite atomizers. …………………………………………… 11
2.1.1 Two step-atomizer designed by Frech. …………………………… 13
2.1.1.1 System description. ……………………………………. 13
2.1.1.2 Features of the Frech system. …………………………. 16
2.1.2 Two-step atomizer by Nagulin. …………………………… …… 17
2.1.3 G. Schlemmer system. …………………………………………. 20
2.1.4 Grinshtein system. ………………………………………………... 22
2.2 THE HIGH TEMPERATURE CHROMATOGRAPHY IN AAS. …… 28
-
VI
2.2.1 High temperature chromatography system proposed by Grinshtein. .. 30
2.2.2 High temperature chromatography with modified
two-step atomizer system. ……..…………………………… 35
3. AIM OF THE WORK. …………………………………………………… 38
4. METHODS AND EXPERIMINTAL. …………………………………… 39
4.1 AAS spectrometer description. ………………………………………. 39
4.2 Graphite materials. …………………………………………………. 39
4.3 Hollow cathode lamps ………………………………………………. 40
4.4 Mass flow controller. 40
4.5 Water and nitric acid. ………………………………………………… 40
4.6 Standard solutions and certified materials. …………………………… 41
4.7 Boronitride. ………………………………………………………… 42
5 RESULTS AND DISCUSSION. ………………………………………… 43
5.1 The new T-shaped graphite furnace. ………………………………… 43
5.2 The short time temperature program. ………………………………. 46
5.3 Optimization of the T-shaped furnace. ……………………………… 49
5.3.1 Optimization of dimensions. ………………………………… 4 9
5.3.2 Optimization of argon gas flow. ………………………………… 49
5.3.3 Optimization of the position
5.3.3.1 Vertical position of furnace neck. …………………………… 51
5.3.3.2 Horizontal position of the furnace neck. …………………… 57
5.4 QUANTITATIVE ANALYSIS WITH THE NON-PYROCOATED
T-SHAPED GRAPHITE FURNACE IN THE HORIZONTAL POSITION. ….. 59
5.4.1 QUANTITATIVE ANALYSIS OF ELEMENTS IN STANDARD
SOLUTIONS. …………………………………………………………. 59
-
VII
5.4.1.1 QUANTITATIVE ANALYSIS OF HIGH VOLATILE ELEMENTS. ….. 59
5.4.1.1.1 Cadmium. ……………………………………………… 59
5.4.1.1.2 Silver. ………………………………………………… 61
5.4.1.1.3 Bismuth. ……………………………………………… 63
5.4.1.2 ANALYSIS OF MIDDLE VOLATILEELEMENTS.
5.4.1.2.1 Calibration curve for cupper. …………………………. 66
5.4.1.2.2 Calibration curve for Manganese. ……………………... 68
5.4.2 ANALYSIS OF ELEMENTS IN URINE. ……………………. 70
5.4.2.1 QUANTITATIVE ANALYSES OF HIGHLY VOLATILE ELEMENTS
5.4.2.1.1 High temperature chromatography in analysis of cadmium. ……… 70
Analysis of cadmium using conventional AAS. ……… 73
5.4.2.1.2 Analysis of Bismuth. ……………………………………….. 75
Analysis of bismuth with conventional Shimadzu system. ……. 77
5.4.2.2 DETERMINATION OF MIDDLE VOLATILE ELEMENS. ………………. 79
5.4.2.2.1 Analysis of Chromium using T-shaped graphite furnace. … 79
Analysis of chromium by using Shimadzu furnace. ………… 82
5.4.2.2.2 Analysis of Manganese. …………………………………... 83
Analysis of manganese using Shimadzu graphite furnace. ..... 85
5.4.2.2.3 Analysis of Aluminium using T-shaped furnace. …………….. 86
Analysis of aluminium using Shimadzu graphite furnace …… 89
5.4.2.2.4 Analysis of copper by non coated T-shaped graphite furnace.….. 90
5.5 ANALYSIS WITH THE PYROCOATED T-SHAPED GRAPHITE
FURNACE. …………………………………………………………………… 92
5.5.1 ANALYSIS OF TRACE ELEMENTS IN URINE. …………………… 92
5.5.1.1 ANALYSIS OF HIGH VOLATILE ELEMENTS. ……………. 92
5.5.1.1.1 Analysis of cadmium. ………………………………… 92
-
VIII
5.5.1.2 ANALYSIS OF MIDDLE VOLATILE ELEMENTS. …….. 94
5.5.1.2.2 Short time and conventional temperature program
in analysis of manganese. …………………………… 95
Analysis of Mn in urine sample using conventional
temperature program. …………………………………….. 95 Analysis of Mn in urine sample using short time temperature
Program. …………………………………………………. 98
5.5.1.2.3 Analysis of of Aluminium. ………………………… 103
5.5.2 ANALYSIS OF TRACE ELEMENTS IN BODY FLUIDS. ……… 104
5.6 SUMMARY OF THE RESULTS. ……………………………………………. 109
5.7 CONCLUSION. ………………………………………………………………. 113
5.8 REFERENCES. ………………………………………………………………. 116
-
IX
LIST FO FIGURES
Figure 1.1 Schematic construction of an atomic absorption spectrometer. ……… 2
Figure 1.2 Relationship between light emission profile and absorption profile. ….. 3
Figure 1.3 A conventional graphite tube with platform. ………………………… . 5
Figure 1.4 Transverse-heated furnace. …………………………………………… 5
Figure 1.5 Heating profiles for the graphite tube walls (A), the inert gas (B),
and the Platform(C). ……………………………………………………….. 6
Figure 2.1 First two-steop atomizer designed by Frech. …………………….. 13
Figure 2.2 Graphite cup and side heated tube with integrated contacts. ……... 14
Figure2.3 Furnace housing with installed cup and tube by
Frech. …………………………………………………………………………… 15
Figure .2.4 Normalized characteristic masses as a function of the gap
distance for Cd and Pb. ……………………………………………………… 15
Figure 2.5 Design of two-step atomizer by Nagulin. ………………………….. 17
Figure 2.6 Space-time structure of the signal of non selective absorption. …… 18
Figure 2.7 Integral signals of non selective absorption in (1) THGA
and (2,3) TSA (2 evaporation; 3, atomization. ………………………………. 19
Figure 2.8 Schlemmer diagram of the two step atomizer. ……………………. 20
Figure 2.9 Atomic and background signals of Cd (a) from bovine liver
and standard solution. …………………………………………………… 21
Figure 2.10 TSAVP designated by Grinstein. …………………………. 22
-
X
Figure 2.11 Schematic diagram of Two-step atomizer with vaporizer
purging by Grinshtein. ……………………………………………………… 23
Figure 2.12 Cahnge of chemical interferences on Pb with Tat in different
Atomizers. …………………………………………………………………….. 25
Figure 2.13 Heated graphite tube with filter. ……………………………………… 30
Figure 2.14 Filter furnace proposed by Grinshtein. ………………………….. 31
Figure 2.15 Delay of Cu signal (A) without filter and (B) with filter. …………… 32
Figure 2.16 Atomic absorption signals with delay of different analytes. …………. 32
Figure 2.17 Van deemter plot for gas chromatography. …………………………. 34
Figure 2.18 Modified Grinstein system. ………………………………… 35
Figure 2.19 Determination of Cd in standard none diluted urine sample
(1) Cd signal; (2) Urine signal. ………………………………………………… 36
Figure 2.20 Effect of protection gas on the transported sample. ……………….. 37
Figure 5.1 Schematic diagram of T-shaped furnace. …………………………… 43
Figure 5.2 Top view of T-shaped furnace. ……………………………………… 44
Figure 5.3 The T-shaped furnace installed in horizontal position inside
the spectrometer. ……………………………………………………… 45
Figure 5.4 Diagram of conventional and short temperature program. ………….. 47
Figure 5.5 High temperature chromatography using T-shaped furnace and
short time temperature program in analysis of Cd in urine. ……………… 48
Figure 5.6 Effect of argon flow on absorption signals of Cd in urine. ………… 50
Figure 5.7 Vertical position of T-shaped furnace with side, front and top
view of the furnace. ……………………………………………………… 51
figure 5.8 Signal for Cd in urine sample using vertical furnace position. ………… 53
Figure 5.9 Signals for chromium determination in acidified none diluted urine.
sample with vertical non coated furnace. …………………………………… 54
-
XI
Figure 5.10 Signals for cupper analysis in standard urine using vertical
non coated T-shaped furnace. …………………………………………….. 55
Figure 5.11 Signals for Bi analysis in non diluted urine sample. ………………… 56
Figure 5.12 Three dimensional diagram for T-shaped furnace in horizontal position. 57
Figure 5.13 Separated Cd signals in non diluted urine sample. …………………. 58
Figure 5.14 Absorption Signals for different standard solutions of Cd. …………. 60
Figure 5.15 Calibration curve for Cd standard solution using non pyrocoated.
T-Shaped graphite furnace. …………………………………………………. 61
Figure 5.16 Calibration curve signals for different concentration of Ag. ……… 62
Figure 5.17 Calibration curve for silver standard solutions. …………………… 63
Figure 5.18 Signals for Bi calibration curve using none pyrocoated
T-shaped furnace. …………………………………………………………… 64
Figure 5.19 calibration curve for standard solution of Bi using none
pyrocoated T-shaped furnace. …………………………………………………. 65
Figure 5.20 Signals for different concentrations of Cu standard solution. ……. 67
Figure 5.21 Calibration curve for Cu determination in standard solutions. ……. 67
Figure 5.22 Signals for Mn calibration curve using none pyrocoated
T-shaped furnace. …………………………………………………………… 69
Figure 5.23 Calibration curve for standard solution of Mn using none
pyrocoated T-shaped furnace. ……………………………………………….. 70
Figure 5.24 Signals for Cd (Red) determinations in standard and spiked
urine sample using the none pyrocoated furnace. …………………………… 72
Figure 5.25 Standard addition method for Cd determination in standard none
diluted urine sample using short temperature program. ……………………. 73
Figure 5.26 Analysis of Cd using original Shimadzu tube. ………………………... 75
-
XII
Figure 5.27 Signals for Bi determination in non diluted urine sample
using non- pyrocoated T-shaped graphite furnace and conventional program…….. 76
Figure 5.28 Calibration curve for determination of Bi in urine using
standard addition method and T-shaped graphite furnace. ………… 77
Figure 5.29 The analyses of Bi in diluted urine standard sample. ………………. 79
Figure 5.30 The calibration curve signals for Cr determination in standard
urine sample. ……………………………………………………… 80
Figure 5.31 Standard addition method for Cr determination in standard
none diluted urine sample. …………………………………………… 81
Figure 5.32 Signal for Cr determination using original shimadzu tube. ………....... 82
Figure 5.33 The calibration curve signals for Mn determination in
standard non diluted urine sample. ……………………………………… 84
Figure 5.34 Standard addition method for determination of manganese in
standard none diluted urine sample. ………………………………… 85
Figure 5.35 Analysis of Mn by Shimadzu furnace using Shimadzu furnace. …….. 86
Figure 5.36 Signal for standard addition method for determination of Al
in urine sample using none pyrocoated T-shaped furnace. ……………… 88
Figure 5.37 Standard addition method for analysis of Al in standard none
diluted urine sample using T-shaped graphite furnace. ………………… 88
Figure 5.38 Signal for Al determination using original shimadzu tube. ………… 89
Figure 5.39 Signals for Cu determination in standard none diluted urine
sample none pyrocoated T-shaped graphite furnace. ……………………..... 91
Figure 5.40 Standard addition method for determination of Cu in urine
using none pyrocoated T-shaped graphite furnace. …………………… 92
Figure 5.42 Signals for analysis of Cu in standard urine sample. ………………… 95
-
XIII
Figure 5.43 Signals for Mn in non diluted non acidified urine sample
Mn in non diluted urine sample. …………………………… 96
Figure 5.44 Mn signals in standard urine sample Seronorm
(Py = 800°C, 1200°C) and 2400°C. ………………………………… 98
Figure 5.45 Signals for Mn in non diluted urine (
-
XIV
LIST OF TABLES
Table 1.1 Temperature program used in conventional AAS systems with
variable pyrolysis and atomization temperatures. …………………… 7
Table 2.1 Characteristic masses obtained by Frech. ……………………….. 16
Table 2.2 Peak absorbance values for Cd at 228.8 nm and background signal
for 10-fold diluted sea water. ……………………………………………… 24
Table 2.3 Pb and Cd determination of spiked urine samples and
recovery values of TSAVP in comparison with the platform atomization……. 26
Table 4.1 Hollow cathode lamps manufacturers, wavelength and current
used for elements under investigation. ………………………………… 40
Table 4.2 Standard materials. ……………………………………………………... 41
Table 4.3 certified materials. ………………………………………………………. 41
Table 5.1 Short time temperature program for Cd in urine sample using
T-shaped furnace and high temperature chromatography approach. ………. 52
Table 5.2 Temperature program for cadmium determination in standard solution
at the resonance line of λ = 228.8 nm. ……………………………….. 59
Table 5.3 Temperature program for silver determination in standard solution
at the resonance line of λ = 328.1 nm. ………………………………… 62
Table 5.4 short temperature program for Bi, λ =223.1 nm determination in standard
solution using T-shaped furnace and continuous flow mode. ………………. 64
-
XV
Table 5.5 Temperature program for Cu determination in standard solution
at the resonance line of λ = 324.7 nm. ……………………………………. 66
Table 5.6 Temperature program for Mn determination in standard solution
at the resonance line of λ = 279.5 nm. …………………………………… 68
Table 5.7 Short time temperature program for quantitative analysis of Cd in
urine sample using non-pyrocoated T-shaped furnace. ……………………. 71
Table 5.8 Temperature program for Bi determination in urine
with non-coated T-shaped furnace and continuous flow analysis. ………… 76
Table 5.9 Temperature program for Cr determination in standard solution
at the resonance line of λ = 357.8 nm. ………………………………… 80
Table 5.10 Temperature program for determination of manganese λ = 279.5 nm
in urine using T-shaped graphite furnace and continuous flow mode. …………. 84
Table 5.11 Temperature program for determination of aluminium λ = 307.3 nm
in urine using T-shaped graphite furnace and continuous flow mode. ………. 87
Table 5.12 Temperature program for determination of copper λ = 324.8 nm
in urine using T-shaped graphite furnace and continuous flow mode. ………. 90
Table 5.13 Short time temperature program for quantitative analysis of Cd λ = 228.8
in urine sample using pyrocoated T-shaped furnace. ………………… 93
Table 5.14 Temperature program for determination of manganese in urine using
pyrocoated T-shaped graphite furnace and continuous flow mode. …… 96
Table 5.15 Short time temperature program used for quantitative analysis of
manganese in urine sample (Biorad Mn < 3.5 µg/l). ………………… 98
Table 5.16 Summary of characteristic masses (Mch) values obtained for
Cd and Bi and comparable values from literature. …………………… 109
-
XVI
Table 5.17 illustrates the Mch values obtained for the Cr and Mn and the
comparable values obtained from the literature survey. ……………… 110
Table 5.18 Characteristic mass, Mch values obtained for Cu and Al and the
comparable values obtained from the literature data. ………………… 111
Table 5.19 Values for trace elements analysis in standard non diluted urine
sample and the certified values. …………………………………… 111
-
XVII
LIST OF ABBREVIATIONS
AAS Atomic absorption spectrometry.
A Absorbance.
BG Background absorbance.
MCh Characteristic mass.
GFAAS Graphite furnace atomic absorption spectrometry.
HGA Heated graphite atomizer.
RSD Relative standard deviation
STPF Stabilized temperature platform.
T Temperature
Tpy Pyrolysis temperature
Tat Atomization temperature
TSA Two-step atomizer.
THGA Transversely heated graphite atomizer.
TSAVP Two-step atomizer with vaporizer purging.
τ Transmittance.
Φo Initial radiant flux.
Φ final radiant flux.
λ Wavelength
-
XVIII
-
XVIII
ABSTRACT
A new T-shaped graphite furnace for atomic absorption spectrometry (AAS) has been
designed for kinetic analysis of trace elements. It is installed in heated graphite atomizer
(GFA-EX7) connected with Shimadzu AAS system (AA6800). The furnace is installed in
two different positions vertically and horizontally between the graphite electrodes. Good
separations of analyte and matrix absorption signal were achieved for many elements with
vertical position with some difficulties in sample injection procedure. Better results with
horizontal position installation of the T-shaped furnace. The quality and performance of the
new design was examined by creation of calibration curves for standard solutions of trace
elements such as Ag, Cd, Bi, Mn, Cr, Al, and Cu..
High temperature gas chromatography approach using T-shaped graphite furnace with short
time temperature program and continuous flow is tested for separation of analyte and
background signals. Analysis of highly interfering matrix such as urine is performed using
this approach. Quantitative analyses of many elements in standard urine sample (seronormTM
Trace Elements Urine LOT NO2525) and urine from Biorad were done. Standard addition
curves for quantifications of high, middle and low volatile elements in that urine sample were
obtained with very small background signals.
More over standard bovine muscle and bovine liver were tested for application of high
temperature chromatography in graphite and short time temperature program using T-shaped
graphite furnace. All cases indicate the analyses of urine sample with high organic and
inorganic matrix components and can be performed with less interferences and better
accuracy. Excellent agreements between measured and certified values were achieved for
analyzed elements in non diluted urine sample.
The working range, characteristic mass and RSD are comparable with the corresponding
analytical data using conventional AAS system.
The new furnace design provides several advantageous features including:
-
XIX
(1) Very low interference effects when analyzing samples with highly interfering matrix such
as non diluted urine, which is not possible to analyze when using the other commercially
available types of graphite furnaces.
(2) Sampling volume up to 100 µl without reducing the intensity of the incident beam.
(3) Short time analysis causing higher sample throughput.
(4) Low contamination risk.
(5) Very simple design.
(6) Application of high temperature chromatography with short temperature program is
possible with highly interfering matrices.
(7) T-shaped furnace can be simply installed in AAS-system.
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
1
1. INTRODUCTION
1.1 TRACE ELEMENT ANALYSIS
Trace element analysis is a process where a sample of some material like, soil, waste
water, drinking water, blood, urine, minerals or chemical compound is analyzed for its
elemental composition. A trace element is the species that present in the sample with
amount less than 0.1%. Species from 0.1- 1% is known as minor constituent, and
components more than 1% treated as major compounds [1].
1.2 SPECTROMETRIC METHODS Spectrometric methods are a large group of instrumental methods that deals with atomic
and molecular spectroscopy. The most widely used spectrometric methods are that which
based on the interactions of the electromagnetic radiations and the matter. Atomic
absorption spectrometry uses the interaction of light with the analyte atoms to measure
the absorption which related to the concentration of the analyte atoms in the gas phase.
1.3 ATOMIC ABSORPTION SPECTROMETRY (AAS)
Atomic absorption spectrometry is a quantitative method for measuring the amount of the
individual analytes in the sample using a special light source at certain wavelength in the
presence of the others. Since samples are usually liquids or solids, the analyte atoms or
ions must converted into free atoms to be measured. The vaporized atoms absorb
ultraviolet or visible light and are transfer into higher energy level. The quantity of light
absorbed by the analyte atoms is proportional to the amount of that analyte in the sample.
1.3.1 INSTRUMENTATION
The general construction of an atomic absorption spectrometer is schematically shown in
figure 1.1. The most important components are; (i) a radiation source, which emits the
spectrum of the analyte element, (ii) an atomizer, in which the atoms of the sample to be
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
2
analyzed are formed, (iii) a monochromator for the spectral dispersion of the radiation,
(iv) a detector permitting measurement of radiation intensity, (v) amplifier and (vi) read
out device that presents a reading.
1.3.2 MEASURING THE ABSORPTION
When the light of radiation source of certain wavelength separated by a monochromator,
and initial intensity is passed through a cell containing atoms in the gas phase in the
electronically ground state, the intensity will decrease to a certain amount caused by
radiation absorption of atoms in the cell. The absorbed radiation then directed to the
detector where the reduced intensity is measured. The amount of light absorbed is
corresponds to the amount of atoms in the cell and finally to the amount of element in the
sample [2].
The usual quantity employed in normal absorption measurements is the radiant flux, Φ.
The LAMBERT- BEER law which relates the absorbance to the concentration can be
used in the following form
Fig .1.1 Schematic construction of an atomic absorption spectrometer
Atomized sample
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
3
A = log Φo / Φ = log 1 / τ 1-1
where A is the absorption, Φo and Φ are the initial and final radiant fluxes, and τ the transmittance.
Atomic absorption spectrometry is a relative method since linear relationship exists
between the concentration of the free atoms in the measurement beam and their
absorbance (Lambert-Beer law). The amount of analyte can be determined using
calibration curve or standard addition method. The calibration curve is only linear within
a certain concentration range of analyte, and then at higher amounts of the analyte
deviation from linearity is observed. It is attributed to the shift in the absorption profile
due the disordered thermal motion of the atoms and various collisions of the analyte
atoms within each other and with atoms or molecules contained in the sample. This shift
is shown in Figure 1.2.
Fig.1.2 Relationship between light emission profile and absorption profile [3].
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
4
1.3.3 GRAPHITE FURNACE
The emission spectrum of analyte element emitted from the radiation source is passed
through an absorption cell in which a portion of incident radiation is absorbed by atoms
produced by thermal dissociation in a flame or a heated tube made of graphite or
tungsten. The most important function of this absorption cell is to produce analyte atoms
in electronic ground state from ions or molecules present in the sample. It is the most
difficult and critical process within the atomic absorption procedure.
The most advanced and widely used highly sensitive system for atomic absorption is
graphite furnace [4]. In this technique, a tube of graphite is located in the compartment of
the spectrometer, with the light beam passing through it. A small volume of sample
solution is placed into the tube, normally through sample injection hole located in the
center of the tube wall. The tube is heated electrically by running a specific heating
program until the analyte present in the sample is dissociated into atoms and absorption
occurs. As atoms are created and transferred out of the tube, the absorbance rises and
falls forming a peak shaped signal. The peak height or integrated peak area is used as
analytical signal for the quantitative determination.
1.3.3.1 SHORT HISTORY OF THE GRAPHITE TUBE
The use of the high-temperature furnace was firstly suggested by the Russian scientist
B.L. L,vov in 1959. It had been developed by many workers. The most popular is the
arrangement of massmann system [5], with graphite tube heated longitudinally between
two electrodes and known as heated graphite atomizer (HGA). The temperature profile
across the length of the tube was not symmetrical and the tube ends were usually at lower
temperature than the centre of the tube. After several years of research, the geometry of
the tube was finally optimized to a length between 20-30 mm and a diameter of 4-6 mm.
Insertion of a platform into the tube was effective method because the platform was
insulated electrically and thermally from the walls of the tube .The platform is heated by
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
5
the radiations from the wall providing the required delay in atomization into the nearly
stable thermal environment [2].
1.3.3.2 STABILIZED TEMPERATURE PLATFORM CONCEPT (STPF)
The insertion of a small platform inside the graphite tube was done by Slavin. The
platform is a flat piece made of pyrolytically-coated graphite and placed near the bottom
of the tube. A during the heating steps the tube walls heat firstly then the platform will be
heated, thus the analyte will be atomized later and the gas inside the tube will reach to
thermal stability before the atomization. This favors free atoms formation, maximizing
sensitivity and producing a constant sensitivity regardless of sample matrix.
Figure 1.3 shows the picture of longitudinally heated graphite tube with platform, and the
transverse-heated graphite atomizer (side heated) is shown in figure 1.4.
Fig.1.3 Conventional graphite tube platform, Fig.1.4 Transverse-heated furnace
The heating profiles of the graphite tube walls, the inert gas, and the platform are shown
in figure 1.5. From the profiles below, the atomization off the wall the sample volatilized
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
6
at a time t1 when the inert gas is still colder; while the sample volatilized with time delay
and the inert gas is stabilized at higher temperature in case of platform t3 [2].
Fig.1.5 Heating profiles for the graphite tube walls (A), the inert gas
(B) and platform(C) [2].
1.3.4 THE TEMPERATURE PROGRAM
The temperature program is composed of several steps, normally four steps, drying,
pyolysis, atomization and cleaning. Each step has its own temperature, holding time,
heating rate and argon flow depending on the type of element wither it is high, middle or
low volatile element and type of sample to be analyzed. Table 1.1 shows conventional
temperature program with variable pyrolysis and atomization temperature depending on
the type of sample and element under investigation.
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
7
Table.1.1 Temperature program used in conventional AAS systems with variable
Pyrolysis and atomization temperatures
1.4 INTERFERENCES IN ATOMIC ABSORPTION SPECTROMETRY
Interferences in graphite furnace atomic absorption spectrometry (GFAAS) can be
classified into three classes, which are chemical, spectral and instrumental interferences.
Interferences mostly arise because of escaping of the analyte off the furnace before
atomization process takes a place. Chemical interferences are result of interactions of the
analyte and the sample matrix components forming molecules. Formation of carbides,
intercalation compounds and other reactions causing incomplete atomization of the
analyte are the main types of chemical interferences.
Spectral interferences can be classified into three sub classes, broad-band background
attenuation, structured interferences, and stray radiation. Broad-band background
attenuation arises from molecular absorption and scattering of the primary source.
Molecular absorption in the ultraviolet and visible regions of the spectrum can be caused
by thermally stable molecules. Scattering interferences arises if some of the matrix
components produce aerosol particles which lead to reduction of the intensity of the
radiation source [6].
Step
T
(°C)
Time
(s)
Mode
Argon flow
(ml/min)
1 120 10 Ramp 500
2 120 20 Hold 500
3 variable 10 Ramp 500
4 variable 20 Hold 500
5 variable variable Hold 10
6 2400 4 Hold 500
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
8
Structured interferences arise when the absorption profile of the undesired element or
molecular structure overlaps the band width of the primary or secondary radiation source.
Stray radiation is radiation with wavelength different from the analyte wavelength and
occurs outside the spectral bandwidth of measurement. Stray radiation or stray light could
be absorbed more or less by the sample causing an error.
Instrumental interferences arises if the instrument is not able to measure the analyte
absorption accurately because of signals rapid and transient, hence the accuracy strongly
depends on the timing associated with the instrument electronics.
In trace element analysis, the most important type of interferences is the chemical
interference which has strong effects on the absorption signal of the analyte atoms. Some
of matrices effects on the analysis are summarized in the next section.
Co-volatilization of trace elements with sodium chloride in the pyrolysis step leads to
loss of sensitivity in the analysis of sea-water samples [8]. Karawowska et al found that
halogenated organic solvents strongly depress iron signal due to the formation of volatile
iron chloride [9]. Another group studied the effect of perchloric acid on the signal of lead
[10]. Fuller discussed the same effect on the signals of thallium [11] and Koirtyohann et
al. [12] showed that the signal of aluminum, gallium and thallium were reduced by 95 %
due to the presence of perchloric acid. Vapor-phase interferences caused also known
effects on the absorption signal in the trace element determination, such kind of
interferences occurs during the atomization stage. Hutton [13] observed strong C2 and
CN molecular bands in the furnace when nitrogen was used as purge gas. L,vov and
Pelievia reported that 30 elements formed monocyanides in presence of nitrogen. L,vov
and Ribzyk reported the absorption spectrum of AlCN for the determination of aluminum
with nitrogen as purge gas, Al2C2 molecular spectrum observed when using argon as
purge gas [14].
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
9
The thermal conditions of the graphite furnace played also of important rule. Frech et al.
showed that the sample volatilized, while the furnace is still heating under non-isothermal
conditions and a part of the analyte vapor phase escaped out of the tube before the
atomization stage [15]. The use of stabilized temperature platform concept leads to more
isothermal conditions and reduces to high extent the volatilization and vapor phase
interferences. Special chemical compounds introduced into the sample solution or the
graphite tube, called chemical modifiers have been found to enhance the sensitivity and
reproducibility by stabilizing the analyte compounds and reducing the matrix effect.
1.5 CHEMICAL MODIFIER
The concept of chemical modifiers was introduced by Ediger in 1973 [16]. Chemical
modifiers can be defined as compounds that are introduced into a graphite atomizer
simultaneously with the test sample, and during the measurement it causes decreasing of
matrix effects. The mechanism of action of most chemical modifiers consists of the
removal of the sample matrix at the pyrolysis stage, this can be attained either by
converting the matrix compounds into volatile compounds or by decreasing the volatility
of the analyte, and gives the possibility to evaporate low volatile matrix components at
temperatures above 1000°C. The use of the unsuitable modifier will not act as a modifier,
but may act as an additional matrix compound and interferes with the analyte absorption.
Chemical modifiers can be classified into the following groups [17]:
(1) Nitric acid and oxalic acids and corresponding ammonium salts ;
(2) Metal nitrates (except platinum-group metals);
(3) Ammonium phosphates;
(4) High melting carbides;
(5) Ascorbic acid, EDTA; and its salts;
(6) Transition metals with higher oxidation states (W+4, Mo+4, Zr+4, etc.).
(7) The universal modifier (Pd/Mg nitrate).
During the last twenty years many workers studied the effectiveness of chemical
modifiers in trace element analysis to reduce the matrix interference. Beinrohr et al. [18]
-
1. INTRODUCTION ------------------------------------------------------------------------------------------------------------
10
studied the effect of ammonium fluoride on the determination of thallium. Shan [19] is
the first one who improved the action of potassium dichromate as matrix modifier for
determination of aluminium in urine. Holcombe studied the function of phosphate
modifier in electrothermal atomizers [20]. The function of ascorbic acid in the
determination of lead by atomic absorption spectrometry is discussed by Chakrabarti in
1989 [21]. Welz studied the effect of mixture of Pd and Ca on the determination of
phosphorous [22]. The effect of various chemical modifiers, including nitrates of
palladium, nickel, magnesium, calcium, lanthanum, europium and aluminium, on the
analytical signal of selenium in a graphite furnace was studied by Docekal in 1991 [23].
Determination of antimony by graphite furnace atomic absorption spectrometry using
five different matrix modifiers, nitric acid, copper, nickel, molybdenum and palladium,
together with L’vov platform was also studied [24]. In 2005 Welz again investigated the
interference of nickel chloride on the determination of copper by electrothermal atomic
absorption spectrometry [25]. Campos studied the function of Iridium as a thermally
deposited permanent modifier on the determination of lead by GFAAS [26]. Pd, Ir and
Rh have been investigated as chemical modifiers for the simultaneous determination of
As, Sb and Se by electrothermal atomic absorption spectrometry [27].
One can observe from the last short survey on the effectiveness of different chemical
modifiers on the analyte under investigation that selection of the chemical modifier
depends on kind of analyte, kind of matrix, temperature program and type of the graphite
furnace used, which in turn means that there are many parameters taken in account for
trace element analysis with GFAAS. Such kind of complex optimization procedure leads
to increase analysis costs by increasing analysis time and instrument and material
consuming.
In the next chapter we will describe many developments of graphite furnace within the
last twenty years in order to reduce the matrix effect in the analysis as much as possible.
Several furnace designs were made and materials rather than graphite such as tungsten is
also studied.
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
11
2. THEORITICAL AND BACKGROUND KNWOLEDGE
2.1 TYPES OF GRAPHITE ATOMIZERS
The first analytical electrothermal atomizer for trace element determinations by atomic
absorption spectrometry was the L, vov furnace [28]. This furnace was a two-step system
and could be used at selected pressures. In this system, sample vaporization and
atomization takes place in two clearly separated units, i.e. graphite electrode on which the
sample is deposited and then vaporized and a heated tube where the sample vapour is
atomized and the analyte concentration measured. This de-coupling of vaporization and
atomization resulted in outstanding performance. At the same time this system was
relatively complex constructed and there fore not suitable for routine work [29].
Due to this separation, this cuvette secured low interference levels and high sensitivity,
but the sample volume was extremely small (1µl) which in contrast led to low sensitivity
and reproducibility. These were the main reasons why the development of commercial
atomizers moved in another direction and the principle of separating vaporization and
atomization zones was neglected for certain period of time.
The Massmann type atomizer combines in the same graphite tube drying, pyrolysis,
vaporization and atomization of the sample. But compared with the L,vov cuvette,
reliability and sensitivity of the analysis decreases, and when analyzing samples with
strongly interfering matrices several chemical interferences occur.
In order to overcome the drawback of the classical electrothermal atomizer, a new oven
design was created, again taking idea of separating vaporization and atomization zone.
One of the most successful solutions was the appearance of graphite furnaces with semi-
separated vaporization and atomization zones in the form of graphite furnaces with
different ballast bodies [30-33]. In this case the sample is injected onto or inside the
ballast body placed in the furnace. The sample will be atomized with some delay if the
ballast in poor contact with the furnace body. STPF atomizer is the most well-known
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
12
version of a ballast body. Atomizers with different ballasts could solve analytical
problems to some extent. More problems could be solved using different matrix
modifiers [34] and transversely-heated graphite furnace with and without end capped
graphite furnaces [35, 36].
The drawback of these furnaces is the limitation of the size and the shape of the ballast
body depending on the size and the shape of the graphite tube itself, because the ballast
body must never block the light beam. This limits the sample volume and hence the
sensitivity of the measurement. Moreover noncomplete separation of vaporization and
atomization zones does not provide the maximum possible atomization degree [37, 38]. A
number of the so called filter furnaces were also developed [31].
In ultra trace determinations of metal impurities, Grinshtein et al. [39] demonstrated the
separation of metal vapour by passing them through porous graphite inserted in the
graphite tube.
Increasing the efficiency of the analyte determination and matrix effect reduction guide
the researchers to think in the complete separation of vaporization and atomization zones,
which could be done by creating two-step atomizer systems. A number of two-step
atomizers have been developed by Frech and coworkers [40-45]. The first two-step
atomizer suggested by Frech in 1983 is shown schematically in figure 2.1. This was
complex design with four electrodes to acts as side heated tube.
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
13
Fig.2.1 The first two-step atomizer designed by Frech in 1983 [45].
During atomization, spatially isothermal conditions were achieved by using side heated
tube. Other two-step atomizers were carried out by different authors. This chapter will
focus on the most important two-step atomizers within the last two centuries.
2.1.1 TWO STEP-ATOMIZER DESIGNED BY FRECH Frech started in developing two-step atomizer systems 25 years ago. In this section we
will describe the most effective system developed by him self in 2000.
2.1.1.1 SYSTEM DESCRIPTION
The two-step atomizer was consisted of two separated parts, one called cup and the other
called tube (see figure 2.2). The system was installed in a modified Perkin-Elmer Analyst
700 with deuterium lamp background correction. The cup was used as vaporizer and a
tube as atomizer. Heating of the graphite cup was controlled by the Analyst 700 and the
tube heated by additional power supply and triggered by AA Win lab software.
Temperature was controlled by an in-house built control unit containing optical feed back
allowed fast heating of the tube to selected temperatures [46].
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
14
Fig.2.2 Graphite cup (right) and side heated tube (left) with integrated contacts [46]
Fig.2.3 shows the furnace housing, formed as a clamshell. In the upper part of the
housing, the contacts of aside-heated graphite tube are clamped between a pair of
graphite cylinders, pressed inside brass tubes, which in turn are tightly pressed into
water-cooled brass blocks.
2.1.1.2 FEATURES OF THE FRECH SYSTEM
For determination of losses of sample vapour through the gap between the cup and the
tube, Cd, Pb, and In standard solution were used. Sensitivity was monitored for cup to
tube distances between 0.20 and 0.95 mm. These results were compared with that which
were carried out for the same elements but with distance between 1.0 and 7.0 mm [47].
The relative signal for 1.0 and 2.0 mm were 1.00 and 1.01 respectively. For the gaps
between 3.0 and 7.0 mm the peak area sensitivity decreased from 0.91 to 0.28. These
results were explained by the presence of an inward convective flow in the gap region.
Decreasing the distance between the cup and tube increase the linear velocity of the
convective flows, and eliminate diffusion losses of vapour through the gap, which
explains the practically unchanged sensitivity for the 1 and 2 mm gaps as shown in fig
2.4.
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
15
Fig.2.3 Furnace housing with installed cup and tube. (1) Water cooled contact block;
(2) Lower brass housing; (3) holder for quartz windows; (4) pneumatically driven
mechanism for opening and closing the furnace [46].
0
0,2
0,4
0,6
0,8
1
1,2
0 0,2 0,4 0,6 0,8 1
cup-tube distance (mm)
Norm
aliz
er c
har.m
ass
PdCd
Fig.2.4 Normalized characteristic masses as a function of the gap distance for Cd and Pb [46].
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
16
The characteristic masses with different volatilities obtained with the two-step atomizer
by Frech and the values for same elements obtained using transversely heated graphite
furnace (THGA) are shown in table 2.1. Better characteristic masses were achieved with
this system as compared with conventional system.
Table 2.1 Characterstic masses (Mch) obtained by Frech TSA and THGA [46]
The observed memory effects for Co as a function of the analyte mass introduced using
the two atomizers for 10 ng of Co introduced, the THGA and two-step atomizer showed
0.5 % and 2 % carry over respectively.
Disadvantages of this system are that the analyte transfer from the atomizer to the
vaporizer is only affected by normal diffusion. Signals of analyte are expected to be very
broad and overlapped on each others in case of real samples. No application of this
system to analysis of real samples has been reported. The loss of the analyte through the
cup-tube distance was also unavoidable.
Element TSA (pg) THGA (pg)
Cd 0.3 1.3
Pb 7.4 30
In 18 80
Co 5.8 17
Al 14 31
Ni 7.6 20
Cr 1.8 7
Ag 1.9 4.5
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
17
2.1.2 TWO-STEP ATOMIZER BY NAGULIN [48]
This work is created in the year 2003 by Nagulin and his research group. They tried to
separate the lower and upper part of the furnace in order to create two-step atomizer in
one furnace. The design is described in figure 2.5 standard tube atomizer (1) was divided
into two parts, A gap was made in the side of the right contact and a mica bad for
insulating the upper and lower part of the furnace was inserted (4).
To apply the current to the tube, the right graphite contact was cut along the furnace and
isolated with mica; such design offers two electrically and thermally isolated parts of the
original tube. A special computer-driven power supply unit allows controlling the heating
of the lower and the upper parts of the graphite furnace.
Fig.2.5 Design of two-step atomizer (1) graphite tube, (2,3) contacts, (4) mica separator,
(5) contact surface, (6) light beam, (7) injection slit [49]
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
18
With this system it was possible to heat the sample firstly in the lower part and
condensate it on the upper colder part of the tube, which could be heated later again up to
the atomization temperature.
Nagulin studied the effect of sodium chloride matrix on the determination of highly
volatile element (cadmium and lead) using the commercially available transverse-heated
graphite furnace and two-step atomizer. He found that, for two step atomizer, the
maximum allowable mass of sodium chloride for which the background at lead and
cadmium lines could be adequately compensated was 17-30 times higher than that for the
commercial atomizer. The effect of sodium chloride could be suppressed because of
sample fractionation and distillation during its evaporation and condensation steps.
Figure 2.6 shows the signal of the nonselective absorption of the matrix at the cadmium
absorption line (228.8nm) for two-step atomizer and transverse-heated atomizer
Fig.2.6. Signal of nonselective absorption of 18µg NaCl at Cd line (228.8nm) in (a) THGA
with platform, (b) TSA [49]
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
19
The advantage of this system that it supports the arrival of the analyte element into a
temperature stabilized atomizer volume (STPF), hence the sensitivity of the analyte
increased and the matrix interference decreased compared with the transverse heated
graphite furnace as shown in figure 2.7.
.
More experiments should be done to improve this system, because there is no information
about both middle and low volatile elements with highly interfering matrices such as
urine, blood and waste water.
The upper part of the graphite furnace is not completely isolated and could be heated by
radiation from the lower part, also because of the double vaporization of the sample,
highly percent of the analyte will be lost through the injection hole of the furnace is
expected. Further more the analyses were carried out along the furnace i.e. in all time
measurements of both analyte and matrix occurs.
Fig.2.7 Integral signals of nonselective absorption in (1) THGA and (2,3)
TSA (2 evaporation; 3, atomization
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
20
2.1.3 G. SCHLEMMER SYSTEM
The two-step atomizer described by Schlemmer [50] consists of the atomizer placed
between two poles of an electromagnetic field providing Zeeman-effect background
correction capability. The tube and cup were heated by independent power supply
enabling the performance of atomic absorption measurements at temporally and spatially
isothermal conditions. This design is shown in figure 2.8.
In the analysis of biological samples by solid sampling electrothermal furnace
background problems occurring due to molecular absorption could be reduced by
charring the sample in the vaporizer cup outside the furnace. The analyte and the
background signals for the determination of cadmium in bovine liver are shown in figure
2.9. As can be seen from this figure, the analyte signal was higher than the background
Fig.2.8 Schlemmer diagram of the two step atomizer; (a) atomizer; (b-d, f-h) contacts
(e) external gas; ( k) vaporizer; ( i ) turning axis; ( j ) horizontal movement piston[50]
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
21
one. Comparable results have been achieved for the determination of magnesium,
potassium, sodium, manganese and zinc in titanium oxide.
This system is practically the same as that designed by Frech [46]. The advantages of this
system were the usage of highly efficient Zeeman-effect background corrector instead of
the deuterium lamp background correction and the possibility of the solid sampling.
.
This system is relatively new, and there is no enough information about the performance
of the system for low volatile elements determinations and influences of highly
interfering matrices such as urine. There is no argon flow through the vaporizer leads to
broad peaks even for Cd (3 seconds) which is in fact a highly volatile elements and gives
Fig.2.9 Atomic and background signals of Cd (a) from bovine liver and standard
solution; (b) Bovine liver in presence of chemical modifier and pyrolysis [50]
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
22
normally a very sharp peak(less than 1.5 seconds). This means that the sample transferred
into the atomizer by normal diffusion, hence high memory effect in the determination of
low volatile elements and very broad peaks of these elements are expected. No more
information for other elements has been published up to now.
2.1.4 GRINSHTEIN SYSTEM
The two-step atomizer with vaporizer purging (TSAVP) consists of a longitudinal heated
graphite atomizer-cuvette and transverse-heated graphite tube-vaporizer placed
horizontally on the side of the cuvette-atomizer as shown in figure 2.10. The atomizer
and vaporizer were heated independently by two power supply units [51]. The length of
the atomizer and the vaporizer were 35 mm and 15-20 mm, respectively. Figure 2.11
shows a schematic diagram of the TSAVP proposed by Grinshtein and the process of
transferring the sample from the vaporizer to the atomizer, and the double vaporization
concept.
Fig.2.10 TSAVP designated by Grinshtein (1) vaporizer; (2) Atomizer; (3) purges gas; (4) outer argon.
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
23
The double vaporization concept based on vaporization of the sample two times, firstly
sample is vaporized, transferred into atomizer and trapped in it then trapped sample re-
evaporized gained in atomizer. This re-evaporization reduced the matrix interferences to
some extent.
The sample vapour reaches the atomizer with the argon flow. Two different modes of
vaporizer purging can be used, continuous mode, or gas stop mode. Mostly sample
volume was 10 µl in the experiments other volumes some times were used [52].
Grinshtein studied the dependence of the analyte and the background signals for sea
water samples on the atomization temperature in three different atomizers using the two-
step atomizer and the heated graphite atomizer (HGA) and conventional Hitachi graphite
furnace system.
HGA showed the highest background signal even with five-fold diluted sample, the
absorbance values were 3.0 for atomization temperatures between 2200 and 2800ºC.
TSAVP showed background absorbance readings between 0.8 and 1.3 which is the
Fig.2.11 Schematic diagram of Two-step atomizer with vaporizer purging; 1
solid sample, 2 atomizer, 3 vaporizer; 4 condensed sample [51].
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
24
lowest signal of the three atomizers with three-fold diluted sample. The values of the
analyte absorbance and the background with different atomizer are shown in the table 2.1
below.
Table 2.2 Peak absorbance values for Cd 228.8 nm and background signal for 10-fold
diluted sea water, sample volume was 10 µl. (3.2 ± 0.3 ng/ml)
The STPF technique was used with HGA and Hitachi atomizers. In addition a Pd
modifier was used with the Hitachi atomizer. In the case of the TSAVP the background
value was small, in the case of Hitachi platform furnace with the matrix modifier very
high. The HGA-700 platform furnace without modifier could not be used because the
background signal was so high that Cd could not be determined.
The effect of the matrix on the absorption signal of Pb was studied by comparing the
absorption signal of 0.5 ng of Pb in distilled water with the same amount of Pb in ten-fold
diluted sea water in different atomizers at different atomization temperatures. The results
are shown in fig 2.12. As can be seen from the figure at higher atomization temperatures
the matrix effect is completely removed using TSAVP since the percent ratio of
absorbance signals of Pb in water and sea water sample is 100%, while the same ratio is
80% when Hitachi and 60% when HGA-700 was used. The argon purge gas in TSAVP
was 20 ml/min.
Atomizer Tat (ºC) Absorbance signal Background signal
TSAVP 1800 0.017 0.15
Hitachi (STPF)+M 2100 0.024 1.6
HGA-700 (STPF) 2000 ----- 2.8
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
25
Analysis of urine sample using stabilized temperature platform furnace STPF, rapidly
heated graphite tube atomizer and chemical modifiers with the sample pretreatment
provides interference free analysis [53-55]. Radzuik and Romanova [56] found that the
STPF techniques cannot be fully free from interferences in the determination of Pb. The
results of recovery measurements with two step atomizers with different ordinary modes
and the conventional platform without modifier for Pb and Cd are shown the in table 2.3
and 2.4.
With TSAVP, the deuterium background corrector was used, therefore diluted urine
samples could only be measured in the ordinary operating mode because the non diluted
urine samples showed higher than 0.6 absorbance values which are out of the working
range of this corrector. For the double vaporization mode the background signal was
reduced to less than 0.3 for Cd and 0.1 for Pb with recovery of about 100% for all
samples under investigation.
Fig.2.12 Cahnge of chemical interferences on Pb with atomization temperature in
different atomizers, Qi is the integrated Absorbance of 0.5 ng Pb in distilled water,
Q the same in ten-fold sea water. 1
0
20
40
60
80
100
120
1800 2000 2200 2400 2600 2800 3000 3200
Atomization temprature (°C)
Q/Q
i %TSAVP
hitachi
HGA
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
26
Table.2.3 Pb and Cd determination of spiked urine samples and recovery values (R) of
TSAVP in comparison with the platform atomization.
Platform atomization TSAVP
Element
Amount added (µg/l)
found conc.(µg/l)
Initial spiked
Rb
( %)
found conc.(µg/)
Initial spiked
Rc ( %)
Cd 1.00
2.00
0.14 ± 0.05 1.5 ± 0.05
2.2 ± 0.05
105
136
0.2 ± 0.01 1.16 ± 0.05
2.2 ± 0.1
96
100
Pb 10.0 2.0 ± 0.5 10.0 ± 0.9 80 2.6 ± 0.17 12.9 ± 0.7 103
a The number of replicates n = 7 . b The samples were threefold and fourfold diluted with water for Cd and Pb respectively. c R= ( conc. in spiked sample– conc. in initial sample/Added conc.)×100%
With Grinshtein system and the double vaporization concept it was possible to measure
the amount of some analytes in presence of highly interfering matrix such as sea water
and urine. The characteristic masses and the recovery of this system are much better
under these working conditions than for the transverse heated atomizer and stabilized
temperature platform systems.
Table 2.3 Determination of Pb and Cd in seronormTM trace element urine reference
sample and recovery values (R) for TSAVP double vaporization in comparison with
platform.
Element Known conc.
(µg/l)
Platform atomization
Found conc. R
(µg/l) ( %)
TSAVP, double vap
Found conc. Rc
(µg/l) ( %)
Cd
Pb
0.35
3.0
0.31 ± 0.07 89
1.9 ± 0.2 63
0.36 ± 0.02 103
3.12 ± 0.2 104
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
27
One of the disadvantages of Grinshtein system is the measurement of the analyte element
along the atomizer tube. Spatially and temporally separation of the analyte and the matrix
is not possible, because the matrix is also measured in the same time and position with
the analyte.
Grinshtein proposed a method [57] using the same principle of the double vaporization to
determine the concentration of Hg in the Russian drinking water using this system,
because the conventional method of Hg determination is complicated and needs a
reduction of the sample firstly by using reducing agent (contamination risk). On the other
hand it is not possible to determine the Hg concentration in drinking water using HGA
and THGA because of the low concentration of Hg, and the relatively high characteristic
masses of Hg with both of them.
The main disadvantages of this system are:
- The complex design of the system (not available commercially).
- The measurement of the analyte concentration along the atomizer tube allows the
matrix also to be measured with the analyte at the same moment.
- The atomizer housing is relatively large and not well protected with argon. This
leads to a shorten life time of the tube.
Other authors tested atomizers manufactured from other materials rather than graphite in
order to overcome the interference problems. Kiyohisa [58] studied the effect of using
tungsten tube for atomic absorption spectrometry. Xiandeng Hou [59] used a tungsten-
coil device for atomic absorption spectrometry, the limit of detection with this device
were a factor of ten from that with the conventional AAS systems. Salido [60]
determined Pb in blood samples using tungsten coil atomizer. Dagmar [61] discussed the
action of tungsten atomizer as modifiers. Reid [62] Built and studied an improved version
of a tungsten filament atomizer for atomic absorption spectrometry. Camero [63] studied
the carbide forming elements using tungsten coil platform.
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
28
Williams [64] introduced commercial tungsten filament Atomizer for analytical atomic
spectrometry. Atomization efficiencies for indium and tin in graphite furnaces from
different atomizer surfaces, i.e. pyrolytically coated platform (PG), palladium modified
pyrolytically coated platform (PdPG) and pyrolytically coated platform, modified with
zirconium carbide (ZrPG), have been studied by Yang [65]. Boronitride platform has
been used in electrothermal AAS for determination of Cd in sees water samples without
chemical modification [66]. Reinaldo studied the use of silica and Nickel atomizer [67].
A direct determination of cadmium by electrothermal atomization atomic absorption
spectrometry with a molybdenum tube atomizer has been investigated [68].
Many other Tungsten devices have been used in analytical atomic spectrometry for
approximately 25 years [69].
2.2 THE HIGH TEMPERATURE CHROMATOGRAPHY IN AAS
As mentioned in previous sections in chapter one and two, many authors tried to
overcome the matrix interference problems accompanied with the trace element analysis
by GFAAS using matrix modifications, addition of ballast bodies inside the furnace,
addition of materials to act as filters, different forms of furnaces (side heated with
platform, graphite rod, tube in tube, graphite capsules, ….etc) and several two-step
atomizer designs. The so-called high temperature chromatography using filter furnace is
also used to overcome some of interference problems. In this section we will discuss
briefly the most important studies in this new area of research.
The use of electrothermal atomizers with graphite filter provides additional possibilities
in reducing matrix effect and metal vapour separation. High temperature chromatography
could be applied within the graphite furnace after installation of graphite filters inside the
graphite tube as mentioned before. The sample is injected in the zone which is separated
from the measuring zone by the filter, heated atomized and the evaporated materials
transferred into the measuring zone by passing it through the filter. The sample
components penetrate the stationary phase of that chromatographic system (filter) with
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
29
different speed depending on their physical interactions with the stationary phase and the
diffusion coefficient of the sample components. If the sample components travel with
different speed they will be detected at different retention time. Thus components of the
sample can be separated.
A number of publications about using of filter furnaces were published within the last
two decades. As example the direct filtration through porous graphite for atomic
absorption analysis of beryllium particulates in air was investigated [70]. Katskov
proposed and discussed the filter furnace as a new atomization concept for electrothermal
atomic absorption spectroscopy [71-77]. In 2003 the same author published the usage of
transverse heated filter atomizer (THFA) in determination of Cd and Pb in urine, and its
analytical performances were investigated using a PerkinElmer SIMAA 6000 atomic
absorption spectrometer [71].
The filter furnace atomizer proposed by Katskov is shown in figure 2.13. The sample is
injected into the filter, then dried, evaporized and atomized within the filter and
transported to the measuring zone. But before measuring the urine sample should be
firstly diluted 5-10 fold. The sensitivity of the determination is not good enough because
the filter inserted inside the furnace reduces the original light path through the furnace as
can be seen in figure 2.13.
The use of filter by Katskov was focused only on the reduction of matrix signal and not
on separation of analyte and matrix from each other. Next section will focus on
separation of analyte and matrices by mean of high temperature chromatography using
graphite filter as stationary phase and measurement of the separated signals using atomic
absorption spectrometry.
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
30
Fig. 2.13 Heated graphite tube with filter developed by Katskov [71]
2.2.1 HIGH TEMPERATURE CHROMATOGRAPHY SYSTEM PROPOSED
BY GRINSHTEIN
In 1997 Grinshtein et al. studied the separation of different metal vapours for atomic
spectroscopy by using high temperature chromatography on graphite and atomic
absorption spectrometry as measuring method [39]. At high temperatures the porous
graphite partition works as a stationary phase of a chromatographic column. The
separation efficiency of graphite filter depends on the temperature of the column and on
the carrier gas flow through the partition. Figure 2.14 indicates the filter furnace
designated by Grinshtein. During the heating of these atomizers the sample vapours enter
the analytical measurement zone after flowing through the porous graphite walls of the
filter with the help of argon gas flow. At high temperatures different atomic vapours
interact with the graphite material by different mechanisms. For example the atoms Cd,
Pb, TI, Ge, Sn, Sb,Bi, Se, Zn are practically inert. Atoms of Fe, Co, Ni, Mn, Cr, Cu, Ti,
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
31
Mo, V and some other metals interact chemically with carbon and form stable or
metastable carbides or solid solutions or intercalation compounds [39].
Figure 2.14 Filter furnace proposed by Grinshtein; A. without filter; B with filter [39]
Figure 2.15 indicates the difference in retention time or appearance time of Cu signal
when using graphite furnaces with and without filter. As can bee seen the retention time
for Cu peak without chromatographic column (filter) is lower than that with graphite
filter. The Cu signal appears after 2 seconds without filter, while 4 seconds in case of
filter. The difference in appearance time could be used in separation of mixtures of metal
vapors by using filter furnace. More elements were studied by the same authors. The
separation of the absorption signals of the elements under investigation are shown in
figure 2.16.
As can be seen from the figure, the high volatile elements appear first with less retention
times, the middle volatile appear later with higher retention times. The separation occurs
due to two main factors; the interactions of the metal atoms with the graphite materials
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
32
and the diffusion coefficient of metals through the graphite filter. More information still
needed in case of body fluid sample analysis.
Fig.2.15 Delay of Cu signal (A) without filter and (B) with filter [39].
Fig.2.16 Atomic absorption signals for metals under investigation showing
different retention times [39].
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
33
Since separation of two or more signals is possible using filter graphite furnace based on
chromatographic principles and AAS detection system. One can calculate some important
terms in chromatography such as separation factor and resolution. The separation factor
(α) between two analytes permit the comparison between two adjacent solutes present in
the same spectrum and given by the following equation [78]
α = tR 2 / tR 1
Where tR1 and tR2 are retention times for first and second components of the separated
signals. The resolution factor between two peaks (R) can also be used for quantification
of peaks separation and is given by the relation below
R = 2(tR 2 - tR 1) / w1 + w2
Where w1 and w2 are peak widths of the two signals. Better resolution could achieve
when the value of R is greater than 1. When R =1.5 it is said that the peaks are baseline
resolved.
Referring to figure 2.16, the separation factor between Zn and Cu is calculated from the
figure (α = 2.8), and the resolution factor R = 1.5, which means that the two peaks are
baseline resolved. The same could be done fore another analytes with good separation
factors.
Another chromatographic property could also be studied for high temperature
chromatography in graphite furnace with AAS detection. The Van Deemter equation in
chromatography relates the variance per unit length of a separation column to the linear
mobile velocity by considering physical, kinetic, and thermodynamic properties of a
separation. These properties include pathways within the column.
H = A + B/ Ū + C·Ū (Van Deemter equation)
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
34
A, B and C may be applied to the graphite furnace as a chromatographic column with
stationary phase (graphite filter). The term A which is known as Eddy diffusion depends
on the different paths used by analyte to pass through the column and it is related to the
uniformity of the graphite filter and it is density. The term B is known as the molecular
diffusion and depends on the diffusion coefficient of the analyte in the mobile phase, the
same behavior also expected because this factor is controlled by the velocity of the carrier
gas through the filter and the term C is known by resistance to mass transfer which
related to number of equilibration of analyte substance between the graphite wall and
filter. The effect of these three factors is summarized in figure 2.17.
Fig.2.17 Van Deemter plot for gas chromatography [79].
The term H in equation is known as height equivalent of theoretical plate from which the
number of theoretical plates within the chromatographic column can be calculated. As
can be seen from figure 2.17 the factor H has optimum value with respect to velocity of
the mobile phase. Temperature of furnace plays also important role in separation
processes.
Speed of the mobile phase
H
-
2. THEORITICAL AND BACKGROUND KNWOLEDGE ------------------------------------------------------------------------------------------------------------
35
2.5.2 HIGH TEMPERATURE CHROMATOGRAPHY WITH MODIFIED
TWO- STEP ATOMIZER [80]
The modified Grinshtein two-step atomizer system is shown if figure 2.18, the analyte is
forced out of the atomizer with the help of the purge gas through a new exit slit (position
No. 4 on the graph) adverse to the inlet slit of the atomizer as shown in figure 2.18. As
can be seen from the figure, the atomized analyte atoms pushed into the atomizer
direction and measured in a smaller measuring zone relative to that with the original
Grinshtein system, then swept out of the atomizer from the new exit slit. With this
modification the interference of the matrix w